import os
import numpy as np
import pandas as pd
from glob import glob
import shutil

# image
import cv2
from skimage.io import imread

# TensorFlow
import tensorflow as tf
from tensorflow.keras import layers, models

# Visualisation libraries

## Text
from colorama import Fore, Back, Style
from IPython.display import Image, display, Markdown, Latex, clear_output

## progressbar
import progressbar

## plotly
from plotly.offline import init_notebook_mode, iplot 
import plotly.graph_objs as go
import plotly.offline as py
from plotly.subplots import make_subplots
import plotly.express as px

## seaborn
import seaborn as sns

## matplotlib
import matplotlib.pyplot as plt
from matplotlib.patches import Ellipse, Polygon
from matplotlib.font_manager import FontProperties
import matplotlib.colors as mcolors
from matplotlib.colors import LinearSegmentedColormap
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
from matplotlib import cm
plt.style.use('seaborn-whitegrid')
plt.rcParams['axes.labelsize'] = 14
plt.rcParams['xtick.labelsize'] = 12
plt.rcParams['ytick.labelsize'] = 12
plt.rcParams['text.color'] = 'k'
%matplotlib inline

import warnings
warnings.filterwarnings("ignore")

Home / Healthcare Analytics and Modeling

CT Images from Cancer Imaging Archive with Contrast and Patient Age

Dataset Description

The dataset is designed to allow for different methods to be tested for examining the trends in CT image data associated with using contrast and patient age. The basic idea is to identify image textures, statistical patterns and features correlating strongly with these traits and possibly build simple tools for automatically classifying these images when they have been misclassified (or finding outliers which could be suspicious cases, bad measurements, or poorly calibrated machines)

Data

The data are a tiny subset of images from the cancer imaging archive. They consist of the middle slice of all CT images taken where valid age, modality, and contrast tags could be found. This results in 475 series from 69 different patients.

TCIA Archive Link - https://wiki.cancerimagingarchive.net/display/Public/TCGA-LUAD

Base = 'ct_medical_images'

TIF_path_dict = {os.path.basename(x): x for x in glob(os.path.join(Base, '*', '*.tif'))}
DCM_path_dict = {os.path.basename(x): x for x in glob(os.path.join(Base, '*', '*.dcm'))}
Data = pd.read_csv(os.path.join(Base,"overview.csv"), index_col = 0)
Data = Data.drop(columns = ['raw_input_path'])
Data['tiff Path'] = Data['tiff_name'].map(TIF_path_dict)
Data['dicom Path'] = Data['dicom_name'].map(DCM_path_dict)
Data['Contrast'] = Data['Contrast'].astype(str)
display(Data.head(10))
display(pd.DataFrame({'Number of Instances':[Data.shape[0]], 'Number of Attributes':[Data.shape[1]]}).style.hide_index())

	Age	Contrast	ContrastTag	id	tiff_name	dicom_name	tiff Path	dicom Path
0	60	True	NONE	0	ID_0000_AGE_0060_CONTRAST_1_CT.tif	ID_0000_AGE_0060_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0000_AGE_0060...	ct_medical_images\dicom_dir\ID_0000_AGE_0060_C...
1	69	True	NONE	1	ID_0001_AGE_0069_CONTRAST_1_CT.tif	ID_0001_AGE_0069_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0001_AGE_0069...	ct_medical_images\dicom_dir\ID_0001_AGE_0069_C...
2	74	True	APPLIED	2	ID_0002_AGE_0074_CONTRAST_1_CT.tif	ID_0002_AGE_0074_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0002_AGE_0074...	ct_medical_images\dicom_dir\ID_0002_AGE_0074_C...
3	75	True	NONE	3	ID_0003_AGE_0075_CONTRAST_1_CT.tif	ID_0003_AGE_0075_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0003_AGE_0075...	ct_medical_images\dicom_dir\ID_0003_AGE_0075_C...
4	56	True	NONE	4	ID_0004_AGE_0056_CONTRAST_1_CT.tif	ID_0004_AGE_0056_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0004_AGE_0056...	ct_medical_images\dicom_dir\ID_0004_AGE_0056_C...
5	48	True	NONE	5	ID_0005_AGE_0048_CONTRAST_1_CT.tif	ID_0005_AGE_0048_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0005_AGE_0048...	ct_medical_images\dicom_dir\ID_0005_AGE_0048_C...
6	75	True	NONE	6	ID_0006_AGE_0075_CONTRAST_1_CT.tif	ID_0006_AGE_0075_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0006_AGE_0075...	ct_medical_images\dicom_dir\ID_0006_AGE_0075_C...
7	61	True	NONE	7	ID_0007_AGE_0061_CONTRAST_1_CT.tif	ID_0007_AGE_0061_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0007_AGE_0061...	ct_medical_images\dicom_dir\ID_0007_AGE_0061_C...
8	51	True	NaN	8	ID_0008_AGE_0051_CONTRAST_1_CT.tif	ID_0008_AGE_0051_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0008_AGE_0051...	ct_medical_images\dicom_dir\ID_0008_AGE_0051_C...
9	48	True	NONE	9	ID_0009_AGE_0048_CONTRAST_1_CT.tif	ID_0009_AGE_0048_CONTRAST_1_CT.dcm	ct_medical_images\tiff_images\ID_0009_AGE_0048...	ct_medical_images\dicom_dir\ID_0009_AGE_0048_C...

Number of Instances	Number of Attributes
100	8

Dataset Distribution

def DatasetDist(Table, Target, PD):
    fig = make_subplots(rows=1, cols=2, horizontal_spacing = 0.02, column_widths=PD['column_widths'],
                        specs=[[{"type": "table"},{"type": "pie"}]])

    # Right
    fig.add_trace(go.Pie(labels=Table[Target].values, values=Table['Count'].values,
                         pull=PD['pull'], textfont=dict(size= PD['textfont']),
                         marker=dict(colors = PD['PieColors'], line=dict(color='black', width=1))), row=1, col=2)
    fig.update_traces(hole=PD['hole'])
    fig.update_layout(height = PD['height'], legend=dict(orientation=PD['legend_orientation']),
                      legend_title_text= PD['legend_title'])
    
    # Left
    T = Table.copy()
    T['Percentage'] = T['Percentage'].map(lambda x: '%%%.2f' % x)
    Temp = []
    for i in T.columns:
        Temp.append(T.loc[:,i].values)

    fig.add_trace(go.Table(header=dict(values = list(Table.columns), line_color='darkslategray',
                                       fill_color= PD['TableColors'][0], align=['center','center'],
                                       font=dict(color='white', size=12), height=25), columnwidth = PD['tablecolumnwidth'],
                           cells=dict(values=Temp, line_color='darkslategray',
                                      fill=dict(color= [PD['TableColors'][1], PD['TableColors'][1]]),
                                  align=['center','center', 'center'], font_size=12, height=20)), 1, 1)

    fig.update_layout(title={'text': '<b>' + Target + '<b>', 'x':PD['title_x'],
                                     'y':PD['title_y'], 'xanchor': 'center', 'yanchor': 'top'})
    fig.show()
    
def DistPlot(Feat, Target, PD, Inp = Data):
    fig = px.histogram(Inp, x = Feat, nbins=PD['nbins'], color= Target, marginal = PD['marginal'],
                       color_discrete_sequence= PD['Bar_Colors'], hover_data=Data.columns)
    fig.update_xaxes(showline=True, linewidth=1, linecolor='Lightgray', mirror=True,
                     zeroline=False, zerolinewidth=1, zerolinecolor='Black',
                     showgrid=False, gridwidth=1, gridcolor='Lightgray')
    fig.update_yaxes(showline=True, linewidth=1, linecolor='Lightgray', mirror=True,
                     zeroline=True, zerolinewidth=1, zerolinecolor='Black',
                     showgrid=True, gridwidth=1, gridcolor='Lightgray')
    # Median
    fig.add_trace(go.Scatter(x= Inp[Feat].median()* np.ones(int(PD['ylims'][1])),
                             y= np.arange(int(PD['ylims'][0]), int(PD['ylims'][1])),
                         name="Median", line=dict(color='RoyalBlue', width=2, dash='dot')))
    # Mean
    fig.add_trace(go.Scatter(x= Inp[Feat].mean()* np.ones(int(PD['ylims'][1])),
                             y= np.arange(int(PD['ylims'][0]), int(PD['ylims'][1])),
                         name="Mean", line=dict(color='Red', width=2, dash='dot')))

    Name = '%s Distribution by %s' % (Target, Feat)
    fig.update_layout(title={'text': '<b>' + PD['Title'] + '<b>', 'x':PD['title_x'],
                             'y':PD['title_y'], 'xanchor': 'center', 'yanchor': 'top'}, yaxis_title='Frequency',
                      plot_bgcolor= 'white', height = PD['height'],
                      legend=dict(orientation=PD['legend_orientation'], x=PD['legend_x'], y=PD['legend_y']),
                      legend_title_text= PD['legend_title'])
    fig.update_traces(marker_line_color= PD['line_color'], marker_line_width=0.5, opacity=1)
    fig['layout']['yaxis'].update(range=PD['ylims'])
    fig.show()

Contrast

Feat = 'Contrast'
Name = 'Contrast'
Table = Data[Feat].value_counts().to_frame('Count').reset_index(drop = False).rename(columns = {'index':Name})
Table['Percentage'] = np.round(100*(Table['Count']/Table['Count'].sum()),2)

Pull = [0 for x in range((len(Table[Name])-1))]
Pull.append(.05)
PD = dict(PieColors = ['RoyalBlue','DeepPink','LawnGreen'], TableColors = ['SlateGray','AliceBlue'], hole = .4,
          column_widths=[0.5, 0.5],textfont = 14, height = 400, tablecolumnwidth = [.1, .05, .08],
          pull = Pull, legend_title = Name, legend_orientation = 'v', title_x = 0.5, title_y = 0.85)
del Pull
DatasetDist(Table, Target = Name, PD = PD)

PD = dict(nbins = 20, Bar_Colors = ['RoyalBlue','DeepPink','LawnGreen'],
          marginal='violin', line_color = 'Black', ylims = [0, 40],
          Title = 'Age and Contrast', title_x = 0.5, title_y = .95, height = 450,
          legend_title = 'Lesion', legend_orientation = 'h', legend_x = 0.01, legend_y = -0.2)
DistPlot('Age', Target = Feat, PD = PD)

Preprocessing

Copying all files into a temporary directory.

def ImgSep(Data, Target, NewDire, Img_Path, file_name, Convert_to_jpg = False):
    # Creating a new directory
    if os.path.exists(NewDire):
        shutil.rmtree(NewDire)

    # Creating sub-directories
    for subfolder in Data[Target].unique().tolist():
        if not os.path.exists(os.path.join(NewDire, subfolder)):
            os.makedirs(os.path.join(NewDire, subfolder))

    Counter = 0
    Progress_Bar = progressbar.ProgressBar(maxval=Data.shape[0],
                                           widgets=[progressbar.Bar('#', '|', '|'), progressbar.Percentage()])
    Progress_Bar.start()
    for _, row in Data.iterrows():
        # from the current dir to a new one
        if Convert_to_jpg:
            cv2.imwrite(os.path.join(NewDire, row[Target], row[file_name].split('.')[0] + '.jpg'),
                        imread(row[Img_Path]), [int(cv2.IMWRITE_JPEG_QUALITY), 100])
        else:
            shutil.copy(row[Img_Path], os.path.join(NewDire, row[Target], row[file_name]))
        Progress_Bar.update(Counter)
        Counter+=1
    Progress_Bar.finish()
    
Path = Base+'_mod'
ImgSep(Data = Data, Target = 'Contrast', NewDire = Path,
       Img_Path = 'tiff Path', file_name = 'tiff_name', Convert_to_jpg = True)

|#########################################################################|100%

Now,

def Path_Tree(PATH, Extension):
    Out = {}
    sep = ' ' * 3
    BACK = {'Black': Back.BLACK, 'Red':Back.RED, 'Green':Back.GREEN, 'Yellow': Back.YELLOW, 'Blue': Back.BLUE,
         'Magenta':Back.MAGENTA, 'Cyan': Back.CYAN}
    title = PATH.split('\\')[-1]
    print(Style.RESET_ALL + Fore.BLUE + Style.NORMAL + '=' * (len(title) +1) + Style.RESET_ALL)
    print(Back.BLACK + Fore.CYAN + Style.NORMAL + title+':'+ Style.RESET_ALL)
    print(Style.RESET_ALL + Fore.BLUE + Style.NORMAL + '=' * (len(title) +1)+ Style.RESET_ALL)
    i = 0
    C = ['Red', 'Green', 'Yellow', 'Blue', 'Magenta', 'Cyan']*len(os.listdir(PATH))
    for entry in os.listdir(PATH):
        if os.path.isdir(PATH):
            print('└──',BACK[C[i]] + Fore.BLACK + Style.NORMAL + entry+':'+ Style.RESET_ALL)
            Sub = os.path.join (PATH, entry)
            List = os.listdir(Sub)
            List = [x for x in List if x.endswith(Extension)]
#             Out[entry] = [os.path.join(Sub, x) for x in List]
            Out[entry] = List
            print(2* sep, Fore.BLUE + Style.NORMAL +
                          '%i %s files:' % (len(List), List[0].split('.')[-1].upper()) + Style.RESET_ALL)
            print(2* sep, ', '.join(List[:5]) + ', ...')
            i+=1
    return Out

_ = Path_Tree(Path, '.jpg')

======================
ct_medical_images_mod:
======================
└── False:
       50 JPG files:
       ID_0050_AGE_0074_CONTRAST_0_CT.jpg, ID_0051_AGE_0063_CONTRAST_0_CT.jpg, ID_0052_AGE_0072_CONTRAST_0_CT.jpg, ID_0053_AGE_0073_CONTRAST_0_CT.jpg, ID_0054_AGE_0082_CONTRAST_0_CT.jpg, ...
└── True:
       50 JPG files:
       ID_0000_AGE_0060_CONTRAST_1_CT.jpg, ID_0001_AGE_0069_CONTRAST_1_CT.jpg, ID_0002_AGE_0074_CONTRAST_1_CT.jpg, ID_0003_AGE_0075_CONTRAST_1_CT.jpg, ID_0004_AGE_0056_CONTRAST_1_CT.jpg, ...

batch_size = 64
(Img_Height, Img_Width) = imread(Data['tiff Path'][0]).shape

train_ds = tf.keras.preprocessing.image_dataset_from_directory(directory= Path,
                                                               validation_split=0.2,
                                                               color_mode = 'grayscale',
                                                               subset="training",
                                                               seed=123, image_size=(Img_Height, Img_Width),
                                                               batch_size=batch_size)
val_ds = tf.keras.preprocessing.image_dataset_from_directory(directory= Path,
                                                             validation_split=0.2,
                                                             subset="validation",
                                                             color_mode = 'grayscale',
                                                             seed=123,
                                                             image_size=(Img_Height, Img_Width),
                                                             batch_size=batch_size)

Found 100 files belonging to 2 classes.
Using 80 files for training.
Found 100 files belonging to 2 classes.
Using 20 files for validation.

fig, ax = plt.subplots(4, 4 , figsize = (16, 16))
ax = ax.ravel()
class_names = train_ds.class_names
for images, labels in train_ds.take(1):
    for i in range(len(ax)):
        _ = ax[i].imshow(images[i].numpy().astype("uint8"), cmap='bone')
        _ = ax[i].set_title('Contrast: %s' % class_names[labels[i]],
                            fontweight='bold', fontsize = 12)
        _ = ax[i].axis("off")
        _ = ax[i].set_aspect(1)

Modeling: Keras Multi-layer Perceptron (MLP) for Image Classifications

A multi-layer perceptron (MLP) is a class of feedforward artificial neural network (ANN). The algorithm at each iteration uses the Cross-Entropy Loss to measure the loss, and then the gradient and the model update is calculated. At the end of this iterative process, we would reach a better level of agreement between test and predicted sets since the error would be lower from that of the first step.

data_augmentation = tf.keras.Sequential([layers.experimental.preprocessing.RandomFlip("horizontal", 
                                                 input_shape=(Img_Height, Img_Width, 1)),
                                         layers.experimental.preprocessing.RandomRotation(0.1),
                                         layers.experimental.preprocessing.RandomZoom(0.1),])

fig, ax = plt.subplots(2, 4 , figsize = (16, 8))
ax = ax.ravel()
class_names = train_ds.class_names
for images, labels in train_ds.take(1):
    for i in range(len(ax)):
        augmented_images = data_augmentation(images)
        _ = ax[i].imshow(augmented_images[0].numpy().astype("uint8"), cmap='bone')
        _ = ax[i].set_title('Augmented Images',
                            fontweight='bold', fontsize = 12)
        _ = ax[i].axis("off")
        _ = ax[i].set_aspect(1)

num_classes = 2
model = models.Sequential([data_augmentation,
                           layers.experimental.preprocessing.Rescaling(1./255),
                           layers.Conv2D(16, 1, padding='same', activation='relu', input_shape=(Img_Height, Img_Width, 1)),
                           layers.MaxPooling2D(),
                           layers.Conv2D(32, 1, padding='same', activation='relu'),
                           layers.MaxPooling2D(),
                           layers.Conv2D(64, 1, padding='same', activation='relu'),
                           layers.MaxPooling2D(),
                           layers.Dropout(0.2),
                           layers.Flatten(),
                           layers.Dense(128, activation='relu'),
                           layers.Dense(num_classes)])
model.summary()
tf.keras.utils.plot_model(model, show_shapes=True, show_layer_names=True, expand_nested = True)

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
sequential (Sequential)      (None, 512, 512, 1)       0         
_________________________________________________________________
rescaling (Rescaling)        (None, 512, 512, 1)       0         
_________________________________________________________________
conv2d (Conv2D)              (None, 512, 512, 16)      32        
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 256, 256, 16)      0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 256, 256, 32)      544       
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 128, 128, 32)      0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 128, 128, 64)      2112      
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 64, 64, 64)        0         
_________________________________________________________________
dropout (Dropout)            (None, 64, 64, 64)        0         
_________________________________________________________________
flatten (Flatten)            (None, 262144)            0         
_________________________________________________________________
dense (Dense)                (None, 128)               33554560  
_________________________________________________________________
dense_1 (Dense)              (None, 2)                 258       
=================================================================
Total params: 33,557,506
Trainable params: 33,557,506
Non-trainable params: 0
_________________________________________________________________

Compiling and fitting the model

# Number of iterations
IT = 51

model.compile(optimizer='adam', loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True), metrics=['accuracy'])

# Training the model
history = model.fit(train_ds, validation_data=val_ds, epochs=IT, verbose = 0)

def Search_List(Key, List): return [s for s in List if Key in s]

Metrics_Names = {'loss':'Loss', 'accuracy':'Accuracy', 'mae':'MAE', 'mse':'MSE', 'recall': 'Recall'}

def Table_modify(df, Metrics_Names = Metrics_Names):
    df = df.rename(columns = Metrics_Names)
    df = df.reindex(sorted(df.columns), axis=1)
    df.insert(loc = 0, column = 'Iteration', value = np.arange(0, df.shape[0]), allow_duplicates=False)
    return df

Validation_Table = Search_List('val_',history.history.keys()) 
Train_Table = list(set( history.history.keys()) - set(Validation_Table))
Validation_Table = pd.DataFrame(np.array([history.history[x] for x in Validation_Table]).T, columns = Validation_Table)
Train_Table = pd.DataFrame(np.array([history.history[x] for x in Train_Table]).T, columns = Train_Table)
Validation_Table.columns = [x.replace('val_','') for x in Validation_Table.columns]

Train_Table = Table_modify(Train_Table)
Validation_Table = Table_modify(Validation_Table)

# Train Set Score
score = model.evaluate(train_ds, batch_size = batch_size, verbose = 0)
score = pd.DataFrame(score, index = model.metrics_names).T
score.index = ['Train Set Score']
# Validation Set Score
Temp = model.evaluate(val_ds, batch_size = batch_size, verbose = 0) 
Temp = pd.DataFrame(Temp, index = model.metrics_names).T
Temp.index = ['Validation Set Score']
score = score.append(Temp)
score.rename(columns= Metrics_Names, inplace = True)
score = score.reindex(sorted(score.columns), axis=1)
display(score.style.set_precision(4))

	Accuracy	Loss
Train Set Score	0.9750	0.1636
Validation Set Score	1.0000	0.0719

def Plot_history(history, PD, Title = False, metrics_names = [x.title() for x in model.metrics_names]):
    fig = make_subplots(rows=1, cols=2, horizontal_spacing = 0.02, column_widths=[0.6, 0.4],
                        specs=[[{"type": "scatter"},{"type": "table"}]])
    # Left
    Colors = ['OrangeRed', 'MidnightBlue', 'purple']
    for j in range(len(metrics_names)):
        fig.add_trace(go.Scatter(x= history['Iteration'].values, y= history[metrics_names[j]].values,
                                 line=dict(color=Colors[j], width= 1.5), name = metrics_names[j]), 1, 1)
    fig.update_layout(legend=dict(x=0, y=1.1, traceorder='reversed', font_size=12),
                  dragmode='select', plot_bgcolor= 'white', height=600, hovermode='closest',
                  legend_orientation='h')
    fig.update_xaxes(range=[history.Iteration.min(), history.Iteration.max()],
                     showgrid=True, gridwidth=1, gridcolor='Lightgray',
                     showline=True, linewidth=1, linecolor='Lightgray', mirror=True, row=1, col=1)
    fig.update_yaxes(range=[0, PD['yLim']], showgrid=True, gridwidth=1, gridcolor='Lightgray',
                     showline=True, linewidth=1, linecolor='Lightgray', mirror=True, row=1, col=1)
    # Right
    if not PD['Table_Rows'] == None:
        ind = np.linspace(0, history.shape[0], PD['Table_Rows'], endpoint = False).round(0).astype(int)
        ind = np.append(ind, history.index[-1])
        history = history[history.index.isin(ind)]
    T = history.copy()
    T[metrics_names] = T[metrics_names].applymap(lambda x: '%.4e' % x)
    Temp = []
    for i in T.columns:
        Temp.append(T.loc[:,i].values)
    TableColors = PD['TableColors']
    fig.add_trace(go.Table(header=dict(values = list(history.columns), line_color=TableColors[0],
                   fill_color=TableColors[0], align=['center','center'], font=dict(color=TableColors[1], size=12), height=25),
                   columnwidth = PD['tablecolumnwidth'], cells=dict(values=Temp, line_color=TableColors[0],
                      fill=dict(color=[TableColors[1], TableColors[1]]),
                      align=['center', 'center'], font_size=12,height=20)), 1, 2)
    if Title != False:
        fig.update_layout(plot_bgcolor= 'white',
                      title={'text': Title, 'x':0.46, 'y':0.94, 'xanchor': 'center', 'yanchor': 'top'},
                          yaxis_title='Frequency')
    fig.show()

PD = dict(Table_Rows = 25, yLim = 1.4, tablecolumnwidth = [0.3, 0.4, 0.4], TableColors = ['Navy','White'])
Plot_history(Train_Table, Title = 'Train Set', PD = PD)
Plot_history(Validation_Table, Title = 'Validation Set', PD = PD)

---

References

1. Kaggle Dataset: CT Medical Images
2. TCGA-LUAD Dataset
3. Albertina, B., Watson, M., Holback, C., Jarosz, R., Kirk, S., Lee, Y., … Lemmerman, J. (2016). Radiology Data from The Cancer Genome Atlas Lung Adenocarcinoma [TCGA-LUAD] collection. The Cancer Imaging Archive.
4. Clark K, Vendt B, Smith K, Freymann J, Kirby J, Koppel P, Moore S, Phillips S, Maffitt D, Pringle M, Tarbox L, Prior F. The Cancer Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository, Journal of Digital Imaging, Volume 26, Number 6, December, 2013, pp 1045-1057.

---